diff --git a/dev/bench/data.js b/dev/bench/data.js index 4775b7ce7..96962e893 100644 --- a/dev/bench/data.js +++ b/dev/bench/data.js @@ -1,5 +1,5 @@ window.BENCHMARK_DATA = { - "lastUpdate": 1686935620080, + "lastUpdate": 1686937246949, "repoUrl": "https://github.com/pomerium/pomerium", "entries": { "Benchmark": [ @@ -51526,6 +51526,42 @@ window.BENCHMARK_DATA = { "extra": "103 times\n2 procs" } ] + }, + { + "commit": { + "author": { + "email": "51246568+kenjenkins@users.noreply.github.com", + "name": "Kenneth Jenkins", + "username": "kenjenkins" + }, + "committer": { + "email": "noreply@github.com", + "name": "GitHub", + "username": "web-flow" + }, + "distinct": true, + "id": "e7703a18919f69d7c3689bb7c51608fb14e173ba", + "message": "add JWT timestamp formatting workaround (#4270)\n\nRego will sometimes serialize integers to JSON with a decimal point and\r\nexponent. I don't completely understand this behavior.\r\n\r\nAdd a workaround to headers.rego to convert the JWT \"iat\" and \"exp\"\r\ntimestamps to a string and back to an integer. This appears to cause\r\nRego to serialize these values as plain integers.\r\n\r\nAdd a unit test to verify this behavior. Also add a unit test that will\r\nfail if the Rego behavior changes, making this workaround unnecessary.", + "timestamp": "2023-06-16T10:36:00-07:00", + "tree_id": "44144644836f6aac845697df597b39eea9427ac9", + "url": "https://github.com/pomerium/pomerium/commit/e7703a18919f69d7c3689bb7c51608fb14e173ba" + }, + "date": 1686937246486, + "tool": "go", + "benches": [ + { + "name": "BenchmarkLoggedInUserAccess", + "value": 15766736, + "unit": "ns/op", + "extra": "78 times\n2 procs" + }, + { + "name": "BenchmarkLoggedOutUserAccess", + "value": 10127560, + "unit": "ns/op", + "extra": "122 times\n2 procs" + } + ] } ] }